On going discussion.
I am a disabled person with multiple impairments that include severe
dyslexia. I have found computers to be very valuable enabling me to participate
with the world of written information. I rely on text to speech facilities along
with voice recognition software. I currently use Microsoft Windows, however I
would like to explore the Linux operating system and its environment. To see if
this could be a valuable alternative to Microsoft's Windows. I'm also very
interested to find out whether the Linux environment can support my dyslexia
I know there was a version of IBM ViaVoice for Linux. However from the inquiries I have made it appears that this is no longer available. This is very disappointing as I am familiar with ViaVoice as I use a both ViaVoice and Dragon dictate, voice recognition software.
So I am often asking myself "will I ever seriously start using the Linux operating system." This is why I decided to post this message in hope that there may be some friendly Linux users who may have some knowledge about available voice recognition software for Linux?
I'm really hoping that there are members of the Linux community out there who can provide some useful information or provide some leads that I could follow up in hope of learning more about voice recognition software for Linux. I would also be interested to learn what other software may be available to support my dyslexia on the Linux environment. Like organisers, mind Mapping, text to speech and any other useful types of software that may support dyslexic people. Perhaps I should mention I am planning to use ubuntu.
If you're able to help me with my quest to use the Linux environment please use the form towards the bottom of this page to provide your feedback.
Many thanks to all Linux users reading this message.
9th October 2006
This message was composed using IBM ViaVoice release 9 USB edition.
Right, oh and by the way, I'm Andy's son in case anyone is wondering. Anyway, right - although I don't know a huge amount about voice recognition software for Linux, I'll tell you basically what little I do.
Firstly, ViaVoice, well here's an article about IBM open sourcing their voice recognition software http://www.theinquirer.net/default.aspx?article=18420 and they did do something, I think they open source the engine of it and made it available for people to use within their own software - and I think there are quite a few sourceforge projects that did do exactly that - however then IBM pulled out and you can't even get hold of the old voice engine anymore - at least not through IBM now. Weird, I've found it for sale at amazon http://www.amazon.com/Via-Voice-Personal-Millennium-Edition/dp/B00004TJ1D/sr=1-3/qid=1168612842/ref=pd_bbs_sr_3/104-5020831-2637554?ie=UTF8&s=software . Unfortunately I can't seem to find any news article that backs up IBM retracting Via Voice support from Linux. Although, here's another article about IBM and it's partial open sourcing of Via Voice http://www.theinquirer.net/default.aspx?article=18617
The next thing, of course, is what other options are available? Well, how about fully open source voice recognition software? Well, there's the CMU Sphinx project - http://cmusphinx.sourceforge.net/ - which looks pretty cool, the various projects contained there include a voice recognition engine written entirely in Java, as well as other systems that are built from various languages combined. The project looks good also because it appears to have Sun driving most the development at the moment, although it's been a while since I've looked into it and it's somewhat confusing to work out exactly what's going on there.
Unfortunately, most article on the web seem to be somewhat outdated. This one is somewhat interesting as it mentions CMU Sphinx, http://software.newsforge.com/article.pl?sid=05/01/13/1445211 An example of this lack of up-to-date information is the Voice Recognition HOWTO guide at the Linux Documentation Project, I have begun making enquiries about updating it and I'll be doing that as soon as I complete some of my back-log of work at my sixth form.
Oh, I remember something else now. I went to LugRadio Live 2006 and asked the guy from Ubuntu who is directing the accessibility side of things there whether there were any chances of a solution in Ubuntu for this. He said - and don't hold me 100% to this as it was a while back - that because there is nothing available for Linux in that area, there wasn't really anything that could be done - however as a short term plan, he could look into getting Via Voice or Dragon Dictate to work better in wine, also - a little more recently, I (as I'm a subscriber to the Ubuntu accessibility mailing list) came across a very rough-hack sounding way of doing something like this which involved running the software inside a virtual machine and using a program to "inject" text from the virtual machine into the Linux apps... something like that..? Anyway, here's the first post of three in that discussion; https://lists.ubuntu.com/archives/ubuntu-accessibility/2006-December/001564.html
Well, I guess that's about it for now. I need to get Sphinx 3 working I think, it'd be really useful to know how good it is. Oh - one other thing, recently searching google for stuff like voice recognition for linux I seem to come across quite a few commercial things come up - but they all appear to be aimed at developers... with money.
Anyway, that's it for now - and I hope this makes at least some sense. If anyone could share anything useful with us it'd be so useful, there really seems to be a lack of up-to-date information on this subject out there.
Thanks for reading this!
Receive on 12 of January 2007
Well I'm somewhat confused! I was rather hoping that their would be someone about with knowledge and perhaps development interest to see a voice recognition environment available for the Linux operating system. However the more one explores this issue the more confusing it seems to get. I can't understand why IBM haven't continued providing an end user voice recognition product.
I've been making some inquiries myself however so far not got any positive the back although I'm still hoping. I was doing a UK search on Google for "voice recognition software for disabled people" and guess what? I found this page on the second page of results, so that has helped me a lot hasn't it?
Anyway I also came across the ICT hub "developing ICT resources for the voluntary and community sector" at http://www.icthub.org.uk and there is an interesting article on open source and Linux, on that site. Well this has given me some ideas to follow up regarding voice recognition software for Linux.
Well, I haven't really got much to add right now. I've asked questions in various places, however not many people seem to have any answers.
I've looked into the CMU Sphinx projects again and, although I really don't have the time, someone has to look into it and I seem to be one of few people interested in this subject. So I'll be trying to get it working and depending on how this goes, updating the Linux Speech-Recognition HOWTO document - which is about 5 years out of date.
To see the trails I've left, follow this link ( https://lists.ubuntu.com/archives/ubuntu-accessibility/2007-February/001812.html ) to see my original post to the Ubuntu accessibility mailing list, and go here ( http://forums.lugradio.org/viewtopic.php?t=3257 ) to see the thread I created at the LUGradio forums.
Receive on 21st of February 2007
have you seen this link about using Wine and Dragon
Received on the 5th of April 2007
Many thanks George for this info. I am going to look into the concept Of Using Wine with DragonDictate.
It is indeed unfortunate that much of the info on speech recognition on linux is outdated. Up to a certain extend this reflects the current state of things. However it's also caused by some really outdated articles remaining on top of the search results, because so many articles have linked to them over the years.
Besides the CMU Sphinx project there are some other promising projects out there though. There is for example Julius/Julian ( http://julius.sourceforge.jp/en_index.php?q=en/index.html ). Julius/Julian is capable of doing command and control (Julian) and dictation (Julius). 'Unfortunately' Julius is a Japanese-based project and even though it has been used for many other languages including English (in the end it can work with probably most languages). However this requires an acoustic model and a language model or a grammar and those are not always freely available.
Some other projects exist too. I would recommend having a look at the forums of the VoxForge project ( http://www.voxforge.org/ ), whore they are trying to solve this problem of the 'missing models'. It's a good source of up to date information.
Received on the 16th of May 2007
The information you posted is certainly of much interest and much appreciated, thank you. Although I've taken some time to make a reply, things have been somewhat hectic over the last few weeks, so this is the first opportunity I had to start looking at the next and voice recognition software.
You are certainly right about VoxForge being a good source for up to date information.
I found a message on the VoxForge forum, leading me to a link to a wiki ubuntu page 4 speech recognition. This looks very promising for the future. Use the following link;
I'm still a little confused as to how best to move forward myself however I certainly have much more to follow up now.
Due to abuse of the form for this discussion "Is there voice recognition software available for Linux?." a decision was made to take the form away. However you can e-mail a contribution for this discussion. Please ensure you put "voice for LINUX", for the subject. Can you also make clear in your e-mail whenever you want your e-mail address published with your contribution on this page.
you are sending me an e-mail that you do not want published on this
page. Please ensure you make this very clear in your e-mail message.
21st of October 2008
Contact info and e-mail details.
Terms and Conditions / Disclaimer.
To ensure you have an up-to-date main menu visit the the contents link.
Hi2u_home| About_Hi2u | ADHD | Animal_Pages | Art_Pages | Awards | Cartoons | Contents | Disability_Issues | Dyslexic_Pages | Education | I'm_Bored | Kids_pages | Links_and_Info | News_&_Events | Search | Support_Groups | Visitors_Input | What's_New
This web page is part of the: hi2u 4 people with hidden
impairments also known as invisible disabilities, web site.
This web site designed by Andy Hayes copyright.